DESCRIPTION: Nonsteroidal anti-inflammatory drugs (NSAIDs) are consumed by tens of millions worldwide. Although they relieve pain and inflammation, we understand poorly their mechanism of action. They also cause serious gastrointestinal and cardiovascular adverse effects and are thought to have caused thousands of deaths. Despite enrolling more than 100,000 patients in randomized trials, we still do not know the NSAID of choice for patients with arthritis and heart disease or if NSAIDs differ in clinical efficacy. Here we propose a paradigm shifting, strategic approach to harvest benefit and manage risk by personalizing therapy with NSAIDs. Data from studies from yeast, mammalian cells, zebrafish, mice and humans will be integrated to develop signaling networks that reflect perturbation by model NSAIDs and that generate hypotheses ultimately addressed by prospective, randomized trials in humans. Our hope is that these iteratively refined models, progressively informed by human data, will lead to algorithms of incremental value to clinicians in the prediction of efficacy and adverse effects. This interdisciplinary strategy will deliver innovative tools and technologies, quantitative models and biomarkers of drug response and if successful will allow more rational prescription of NSAIDs to minimize risk and maximize benefit to individuals, creating a novel paradigm for the development and approval of drugs, the design of randomized trials and the treatment of chronic disease.